[PYTHON DOC-SIG] Re: [PROGENV-SIG] Project proposal discussion document

[Jim Fulton]

Fred L. Drake wrote:
> 
>   I am advancing a proposal for a project which would benefit from the
> input of the members of both doc-sig and progenv-sig.  I have prepared a
> discussion document describing the project in general terms in more detail
> than I present in this message, but a brief summary is appropriate.
> 
>   My intention is to produce tools which can analyze Python source code,
> potentially as it is imported into a running Python process, and perform
> sufficient analysis to create a representation of the object hierarchy
> with as much documentation as can be derived from the source text and
> documentation strings as possible. The data structure created by the
> analysis should allow for programmatic examination as part of an
> interactive development or debugging environment or for the generation of
> static documentation of a collection of modules.
> 
>   My hope is that this project can be a collaborative project of,
> primarily, the progenv SIG with the doc SIG providing critical advisory
> input and evaluation. As described above, the work appears to be of
> sufficient general interest to warrent the involvement of many
> contributors. At the same time, the work promises enough modularity that
> individuals need not invest huge amounts of time to provide a meaningful
> contribution. The most significant investment to be made before
> applications may be created within the project framework involves reaching
> a consensus on the definable stages of the project and precise definitions
> of the underlying data architecture.
> 
>   The discussion document is available via the world wide web at:
> 
>                 http://weyr.cnri.reston.va.us/progenv/
> 
>   Feedback is invited via the SIG mailing lists.  I will be attending the
> Python conference next week and hope to bring this proposal up at the
> doc-sig and progenv-sig working groups.

This sounds like an interesting discussion topic for next week.  I look forward
to it.

I have a couple of questions:

  1. I couldn't tell from your note or referenced web pages
     what problems you are trying to solve.  Could you elaborate?

  2. Why is it necessary to parse sources or manipulate parse trees?
     This limits the scope to python objects, but browsing C, ILU and
     other types of objects seems just as interesting to me.  Wouldn't
     it be better to define and implement necessary object protocols
     for discovery and browsing?  What problems require source analysis?

Jim

-- 
Jim Fulton         Digital Creations
jim@digicool.com   540.371.6909

=================
DOC-SIG  - SIG for the Python Documentation Project

send messages to: doc-sig@python.org
administrivia to: doc-sig-request@python.org
=================